Top AI Agents Built to Catch Malicious Code Can Be Tricked Into Running It

Ask an AI coding agent to scan open-source code for security holes, and it might run the attacker's code on your own machine instead. That is the finding in a proof-of-concept published Wednesday by the AI Now Institute, an attack it calls "Friendly Fire. " It works against Anthropic's Claude Cod...

CVE-2026-14898

The OpenAI Codex desktop app for macOS rendered remote images from Markdown in model responses. An attacker who could place an indirect prompt injection in content processed by Codex, such as a connected-tool result or another untrusted source, could induce the model to construct a remote image U...

CVE-2026-14898

The OpenAI Codex desktop app for macOS rendered remote images from Markdown in model responses. An attacker who could place an indirect prompt injection in content processed by Codex, such as a connected-tool result or another untrusted source, could induce the model to construct a remote image U...

CVE-2026-14898

The OpenAI Codex desktop app for macOS rendered remote images from Markdown in model responses. An attacker who could place an indirect prompt injection in content processed by Codex, such as a connected-tool result or another untrusted source, could induce the model to construct a remote image U...

CVE-2026-14898

The OpenAI Codex desktop app for macOS rendered remote images from Markdown in model responses. An attacker who could place an indirect prompt injection in content processed by Codex, such as a connected-tool result or another untrusted source, could induce the model to construct a remote image U...

CVE-2026-14898: Exposure of Sensitive Information to an Unauthorized Actor

The OpenAI Codex desktop app for macOS rendered remote images from Markdown in model responses. An attacker who could place an indirect prompt injection in content processed by Codex, such as a connected-tool result or another untrusted source, could induce the model to construct a remote image U...

Apple Patches 30+ iOS, macOS, Safari Flaws, Including AI-Discovered WebKit Bugs

Apple on Monday released security updates for iOS, macOS, and the Safari web browser to address over three dozen flaws, including four vulnerabilities in WebKit that were discovered using artificial intelligence AI tools like Anthropic Claude and OpenAI Codex Security. The WebKit vulnerabilities...

Maestro 0.15.4

Maestro is a cross-platform desktop app for orchestrating your fleet of AI agents and projects. It's a high-velocity solution for hackers who are juggling multiple projects in parallel. Designed for power users who live on the keyboard and rarely touch the mouse. Collaborate with AI to create...

(0Day) OpenAI Codex Sandbox Escape Vulnerability

This vulnerability allows remote attackers to bypass the sandbox on affected installations of OpenAI Codex. User interaction is required to exploit this vulnerability in that the target must use Codex to process a repository containing malicious JavaScript. The specific flaw exists within the...

CVE-2025-61260

A vulnerability was identified in OpenAI Codex CLI v0.23.0 and before that enables code execution through malicious MCP Model Context Protocol configuration files. The attack is triggered when a user runs the codex command inside a malicious or compromised repository. Codex automatically loads...

@4via6/relay (>=1.0.0, <=1.1.3), @circleci/agents (>=1.0.6, <=2.3.1) +6 more potentially affected by CVE-2025-61260 via @openai/codex (>=0.1.2504211509, <=0.20.0)

@openai/codex NPM version =0.1.2504211509, =1.0.0, =1.0.6, =0.0.1, =0.23.0, =1.0.0, =0.0.1, =0.1.0, =1.0.0, =1.0.2 Source cves: CVE-2025-61260 Source advisory: SNYK:JS-OPENAICODEX-16111764...

OpenAI Codex CLI enables code execution through malicious MCP (Model Context Protocol) configuration files

A vulnerability was identified in OpenAI Codex CLI v0.23.0 and before that enables code execution through malicious MCP Model Context Protocol configuration files. The attack is triggered when a user runs the codex command inside a malicious or compromised repository. Codex automatically loads...

EUVD-2025-209435

A vulnerability was identified in OpenAI Codex CLI v0.23.0 and before that enables code execution through malicious MCP Model Context Protocol configuration files. The attack is triggered when a user runs the codex command inside a malicious or compromised repository. Codex automatically loads...

GHSA-XRXF-JGV3-QMRM OpenAI Codex CLI enables code execution through malicious MCP (Model Context Protocol) configuration files

A vulnerability was identified in OpenAI Codex CLI v0.23.0 and before that enables code execution through malicious MCP Model Context Protocol configuration files. The attack is triggered when a user runs the codex command inside a malicious or compromised repository. Codex automatically loads...

Arbitrary Code Injection

Overview @openai/codex is a OpenAI Codex CLI Lightweight coding agent that runs in your terminal Affected versions of this package are vulnerable to Arbitrary Code Injection via the automatic loading of .env and .codex/config.toml files when executing the CLI in a compromised repository. An...

@4via6/relay (>=1.0.0, <=1.1.3), @circleci/agents (>=1.0.6, <=2.3.1) +6 more potentially affected by CVE-2025-61260 via @openai/codex (>=0.1.2504211509, <=0.20.0)

@openai/codex NPM version =0.1.2504211509, =1.0.0, =1.0.6, =0.0.1, =0.23.0, =1.0.0, =0.0.1, =0.1.0, =1.0.0, =1.0.2 Source cves: CVE-2025-61260 Source advisory: OSV:GHSA-XRXF-JGV3-QMRM...

CVE-2025-61260

A vulnerability was identified in OpenAI Codex CLI v0.23.0 and before that enables code execution through malicious MCP Model Context Protocol configuration files. The attack is triggered when a user runs the codex command inside a malicious or compromised repository. Codex automatically loads...

OpenAI Codex CLI 安全漏洞

OpenAI Codex CLI is a lightweight coding agent software developed by OpenAI and run in the terminal. OpenAI Codex CLI versions prior to v0.23.0 have security vulnerabilities; these vulnerabilities stem from automatically loading malicious MCP configuration files, which may allow arbitrary code to...

CVE-2025-61260: Improper Control of Generation of Code

A vulnerability was identified in OpenAI Codex CLI v0.23.0 and before that enables code execution through malicious MCP Model Context Protocol configuration files. The attack is triggered when a user runs the codex command inside a malicious or compromised repository. Codex automatically loads...

CVE-2025-61260

A vulnerability was identified in OpenAI Codex CLI v0.23.0 and before that enables code execution through malicious MCP Model Context Protocol configuration files. The attack is triggered when a user runs the codex command inside a malicious or compromised repository. Codex automatically loads...